Bayesian regularized quantile regression
Authors:
- Qing LiQLQing Li
- Nan LinNLNan Lin
- Ruibin XiRXRuibin Xi
Open Access
- 1 September 2010
- journal article
- research article
- Published by Institute of Mathematical Statistics in Bayesian Analysis
- Vol. 5 (3) , 533-556
Abstract
Regularization, e.g. lasso, has been shown to be effective in quantile regression in improving the prediction accuracy (Li and Zhu, 2008; Wu and Liu, 2009). This paper studies regularization in quantile regressions from a Bayesian perspective. By proposing a hierarchical model framework, we give a generic treatment to a set of regularization approaches, including lasso, group lasso and elastic net penalties. Gibbs samplers are derived for all cases. This is the first work to discuss regularized quantile regression with the group lasso penalty and the elastic net penalty. Both simulated and real data examples show that Bayesian regularized quantile regression methods often outperform quantile regression without regularization and their non-Bayesian counterparts with regularization.Keywords
This publication has 26 references indexed in Scilit:
- Flexible Bayesian quantile regression for independent and clustered dataBiostatistics, 2009
- Bayesian Semiparametric Modelling in Quantile RegressionScandinavian Journal of Statistics, 2009
- Quantile pyramids for Bayesian nonparametricsThe Annals of Statistics, 2009
- Composite quantile regression and the oracle model selection theoryThe Annals of Statistics, 2008
- The Bayesian LassoJournal of the American Statistical Association, 2008
- L1-Norm Quantile RegressionJournal of Computational and Graphical Statistics, 2008
- The Group Lasso for Logistic RegressionJournal of the Royal Statistical Society Series B: Statistical Methodology, 2008
- Regularization and Variable Selection Via the Elastic NetJournal of the Royal Statistical Society Series B: Statistical Methodology, 2005
- Sparsity and Smoothness Via the Fused LassoJournal of the Royal Statistical Society Series B: Statistical Methodology, 2004
- Modeling Regression Error With a Mixture of Polya TreesJournal of the American Statistical Association, 2002